Samenvatting
Summary All Articles Collab & Advanced Topics Exam
This is a summary of all the mandatory articles from the cource manual of Collaboratoin & Advanced topics which are part of the exam.
[Meer zien]Voorbeeld 4 van de 74 pagina's
Voorbeeld 4 van de 74 pagina's
In winkelwagengesponsord bericht van onze partner
1 beoordeling
Door: rtmwintjens • 6 maanden geleden
Capacious
Voorbeeld van de inhoud
TOPIC 1; Smart assetsArticle 1 - Tiddens, W. W., Braaksma, A. J. J., & Tinga, T. (2015). The
adoption of prognostic technologies in maintenance decision making:
a multiple case study.
good overview of condition-based maintenance methods and observations from practice
Abstract
This study examines the adoption and use of prognostic maintenance technologies in helping asset
owners make optimal decisions about maintaining or extending the life of physical assets. Despite
various techniques being available, only a small number of companies have actually implemented
them. The study investigates the reasons for this and compares it to existing literature. It presents a
framework for implementing prognostic technologies for maintenance decisions, based on literature
postulates. The study's multiple-case analysis in Dutch industries reveals challenges such as
identifying relevant parameters, translating data into decision support, and determining the
appropriate prognostic technology route.
Introduction
Prognostic techniques to assists asset owners can be used to reduce business and safety risks caused
by unexpected failures of critical systems and reduce lifecycle costs. However, there is a gap between
potential and realized benefits leading to being only ‘satisfactory’.
- Prognostic techniques → predict future state of systems.
- Diagnostic techniques → retrospective by nature: identify and quantify the damage occurred
to determine the cause-and-effect relationships (for root causes)
- Detection → closely related to diagnostics and aims to detect anomalies. Binary in nature:
either healthy or faulty system.
Much research is conducted, but many
prognostics and health management
methods are introduced and applied to
solve specific problems without much
explanation or documentation given as to
how or why these methods have been
used. The basic deliberation of how to
conduct an advanced maintenance
analysis can be explained via different
pathways. As seen on the picture. The
methods consist out of different steps:
,Analyses
In this paper a framework is conducted to guide users of prognostic technologies through the
different steps and to aid business purposes rather than only using for technical evaluations.
Six postulates are introduced in three paragraphs which are devoted to three consecutive steps of the
proposed framework.
Step 1: Monitoring and data gathering
- Main categories asset data:
o Condition monitoring data: Collected via condition and health monitoring sensors,
usage and load monitoring systems.
o Event data is gathered from historical record and ERP systems.
- Postulate (=assumed based on reasoning) 1: The collected data is often not useful for
advanced maintenance analyses
This is often assumed in literature. However, in the real world, data collected are not
necessarily in a readily usable form due to issues such as missing data, redundant data, noise
or sensor degradation problems
- Postulate 2: The selection of parameters to monitor is not well motivated
Sensor placement and selection requires knowledge about system’s failure mechanisms and
governing loads. However, it is often discovered after data collection (during interpretation)
that essential quantities are missing and non-relevant parameters have been monitored.
Step 2: The selection of prognostics analysis type and analysis (Advanced
maintenance analyses)
- Two categorizations:
The initial model with 4 categories is extended with an experience-based route by
considering the difference between methods using historical records and those only using
expert knowledge/experience of people operating and maintaining the equipment as they
are the best source of information.
o 0. Experience based predictions: based on knowledge and previous experience
outside (OEM) or inside the company. Sometimes supported by little or scattered
data.
o 1. Reliability Statistics prediction analyses: based on historical (failure records of
comparable equipment without considering the specific (usage) differences.
Describes population-based probabilities. Estimation of an average component
operating under historically average conditions. Based on normal distribution
o 2. Stressor based predictions: based on historical records supplemented with
stressor data, e.g. temperature or humidity, to include environmental and
operational variances. Results in expected lifetime of an average system in specific
environment. Based on extrapolation (=assumption current trends will continue) of a
general path derived from physical models, built in tests, or operating history.
o 3. Degradation based predictions: based on the extrapolation of a general path of a
prognostic parameter (=any variable associated with a subsequent outcome) a
degradation measure (=situations where things break down slowly over time due to
damage piling up), to a failure threshold. The prognostic parameter is inferred from
sensor readings. The prediction includes the current state of degradation and results
in an expected lifetime of a specific system in a specific environment. It also
, measures symptoms of incipient failure e.g. raise in temperature or vibration (root-
cause).
o 4. Mechanism based predictions: based on direct sensing of the critical failure
mechanisms of individual components. It results in an expected lifetime of a specific
system in specified conditions. The prognostic parameter is calculated with a physical
model of the degradation mechanism. The model uses the sensed variation of loads
or usage as input.
- Second categorization:
Classifies maturity levels on type of input data, knowledge or physical model.
o Data: Relies on assumption that the characteristics of the data will remain relatively
unchanged unless malfunction occurs.
+ Strength: ability to transform high-dimensional noise data in two lower
dimensional information.
- Drawback: successfulness highly dependent on quality and quantity of data
o Physical models: use mathematical models. Behavior failure mode is quantitatively
characterized using physical laws and are especially useful for predicting system
response to new loading conditions or new system configurations.
Require an accurate mathematical model.
o Knowledge based models: accumulate experience from subject matter to form rules
to apply that knowledge.
Requires high degree of completeness to be useful.
High amounts of in- and outputs make them very complex.
Postulate 3: Higher maturity levels of maintenance analyses result in higher value analyses
- Mechanism-based maintenance analyses, though more challenging to develop, offer higher
value for decision-makers; however, metrics for justifying investments in advanced
maintenance, like Condition-Based-Maintenance, may show technical improvements without
ensuring increased production line efficiency.
Postulate 4: The predictive performance of the prognostic systems improves in time, they are evolving
systems
- Prognostic systems can be validated and improved during their lifetime because more data is
collected during its utilization. Especially knowledge-based systems should be updated.
Postulate 5: The selection of the type of advanced maintenance analysis is not well motivated
- In literature many maintenances analysis is developed and proposed, however, most are
application or equipment specific. A clear way to design and implement does not exist.
Moreover many are introduced and applied without much explanation or documentation
given as to how or why these methods have been selected.
, Step 3: Technical results and life cycle management decision making
support
Human decisions are often not reliable/accurate when dealing with complex systems. In figure 1, 3a
should be used to improve 3b. Decision support systems (DSS) can be used in step 3b to aid the
decision-making process.
- Model based > human decision making.
Typical DSS consists of a databank, a model (translate data) and dialog (generate insight)
Postulate 6: The quality level of current analyses is not sufficient to improve maintenance decisions
- For effective decision-making using analyses, the forecasted remaining useful lifetime should
exceed the time needed for preventive actions, known as the flexibility phenomenon;
o however, systems with high prognostic distance may be costly and less accurate,
o while a lack of prognostic distance hinders scheduling preventive maintenance
- challenges also arise from limited adoption of decision support systems due to data
availability and integration issues with operational systems.
Case study results
Multiple case study with 4 companies
Postulate 1: The collected data is often not useful for advanced maintenance analyses
Mature companies tend to have more structured and accessible data, while less mature ones
face challenges with fragmented and difficult-to-access data. For mature companies, data
may be automatically sent to integrated data systems and are well accessible for analysis.
Also the quality of data is higher.
So, for higher mature companies the postulate is rejected. For companies with lower maturity
the postulate is accepted
Postulate 2: The selection of parameters to monitor is not well motivated
Companies often measure numerous parameters without clear motivation, relying on trial
and error to determine the effectiveness of the selected parameters. The length of this
process seems dependent on the knowledge level of the company. Success of the outcomes
are possible suboptimal
So, postulate can be accepted.
Postulate 3: Higher maturity levels of maintenance analyses result in higher value analyses
Difficulty in expressing the value of maintenance analyses was noted, with no clear
correlation between the difficulty or maturity of analyses and their perceived value. The
reason for companies to invest in maintenance analyses is often based on an incentive to
lower cost, but it is hard to create a rigid business case since the outcomes of the analyses
are unknown on forehand. Companies therefore rely on the trust of management to invest in
these techniques.
So, postulate rejected: higher mature analyses does not always result in higher value.
Postulate 4: The predictive performance of the prognostic systems improves in time, they are
evolving systems
In the early lifetime of the monitoring system, prognostic techniques are used to learn about
the system instead of directly predicting failures. The knowledge base for understanding the
Voordelen van het kopen van samenvattingen bij Stuvia op een rij:
Verzekerd van kwaliteit door reviews
Stuvia-klanten hebben meer dan 700.000 samenvattingen beoordeeld. Zo weet je zeker dat je de beste documenten koopt!
Snel en makkelijk kopen
Je betaalt supersnel en eenmalig met iDeal, creditcard of Stuvia-tegoed voor de samenvatting. Zonder lidmaatschap.
Focus op de essentie
Samenvattingen worden geschreven voor en door anderen. Daarom zijn de samenvattingen altijd betrouwbaar en actueel. Zo kom je snel tot de kern!
Veelgestelde vragen
Wat krijg ik als ik dit document koop?
Je krijgt een PDF, die direct beschikbaar is na je aankoop. Het gekochte document is altijd, overal en oneindig toegankelijk via je profiel.
Tevredenheidsgarantie: hoe werkt dat?
Onze tevredenheidsgarantie zorgt ervoor dat je altijd een studiedocument vindt dat goed bij je past. Je vult een formulier in en onze klantenservice regelt de rest.
Van wie koop ik deze samenvatting?
Stuvia is een marktplaats, je koop dit document dus niet van ons, maar van verkoper louran. Stuvia faciliteert de betaling aan de verkoper.
Zit ik meteen vast aan een abonnement?
Nee, je koopt alleen deze samenvatting voor €7,00. Je zit daarna nergens aan vast.
Is Stuvia te vertrouwen?
4,6 sterren op Google & Trustpilot (+1000 reviews)
Afgelopen 30 dagen zijn er 53340 samenvattingen verkocht
Opgericht in 2010, al 14 jaar dé plek om samenvattingen te kopen